Operations management 1e cachon 1

www.downloadslide.net www.downloadslide.net Operations Management Final PDF to printer www.downloadslide.net McGraw-Hill Education Operations and Decision Sciences Operations Management Beckman and Rosenfield Operations ­Strategy: Competing in the 21st Century First Edition Benton Purchasing and Supply Chain Management Third Edition Bowersox, Closs, and Cooper Supply Chain Logistics Management Fifth Edition Brown and Hyer Managing Projects: A Team-Based Approach Second Edition Burt, Petcavage, and Pinkerton Supply Management Ninth Edition Cachon and Terwiesch Operations Management First Edition Cachon and Terwiesch Matching Supply with Demand: An ­Introduction to Operations Management Third Edition Finch Interactive Models for Operations and ­Supply Chain Management First Edition Fitzsimmons and Fitzsimmons Service Management: Operations, Strategy, Information Technology Eighth Edition Gehrlein Operations Management Cases First Edition Harrison and Samson Technology Management First Edition Hayen SAP R/3 Enterprise Software: An Introduction First Edition Hill Manufacturing Strategy: Text & Cases Third Edition Hopp Supply Chain Science First Edition Hopp and Spearman Factory Physics Third Edition Jacobs, Berry, Whybark, and Vollmann Manufacturing Planning & Control for ­Supply Chain Management Sixth Edition Jacobs and Chase Operations and Supply Chain Management Thirteenth Edition Jacobs and Chase Operations and Supply Chain Management: The Core Fourth Edition Jacobs and Whybark Why ERP? First Edition Johnson, Leenders, and Flynn Purchasing and Supply Management Fifteenth Edition Larson and Gray Project Management: The Managerial ­Process Sixth Edition Schroeder, Goldstein, and Rungtusanatham Operations Management: Contemporary Concepts and Cases Sixth Edition Simchi-Levi, Kaminsky, and Simchi-Levi Designing and Managing the Supply Chain: Concepts, Strategies, Case Studies Third Edition Sterman Business Dynamics: Systems Thinking and Modeling for Complex World First Edition Stevenson Operations Management Twelfth Edition Swink, Melnyk, Cooper, and Hartley Managing Operations Across the Supply Chain Third Edition Thomke Managing Product and Service ­Development: Text and Cases First Edition Ulrich and Eppinger Product Design and Development Sixth Edition Zipkin Foundations of Inventory Management First Edition Quantitative Methods and Management Science Hillier and Hillier Introduction to Management Science: A Modeling and Case Studies Approach with Spreadsheets Fifth Edition cac42205_fm_i-xviii.indd ii Stevenson and Ozgur Introduction to Management Science with Spreadsheets First Edition 04/20/16 07:21 AM www.downloadslide.net Operations Management Gérard Cachon The Wharton School, University of Pennsylvania Christian Terwiesch The Wharton School, University of Pennsylvania www.downloadslide.net OPERATIONS MANAGEMENT Published by McGraw-Hill Education, Penn Plaza, New York, NY 10121 Copyright © 2017 by McGrawHill Education All rights reserved Printed in the United States of America No part of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval system, without the prior written consent of McGraw-Hill Education, including, but not limited to, in any network or other electronic storage or transmission, or broadcast for distance learning Some ancillaries, including electronic and print components, may not be available to customers outside the United States This book is printed on acid-free paper DOW/DOW ISBN 978-1-259-14220-8 MHID 1-259-14220-5 Senior Vice President, Products & Markets: Kurt L Strand Vice President, General Manager, Products & Markets: Marty Lange Vice President, Content Design & Delivery: Kimberly Meriwether David Managing Director: James Heine Brand Manager: Dolly Womack Director, Product Development: Rose Koos Lead Product Developer: Michele Janicek Product Developer: Christina Holt Marketing Manager: Britney Hermsen Director of Digital Content Development: Douglas Ruby Digital Product Analyst: Kevin Shanahan Director, Content Design & Delivery: Linda Avenarius Program Manager: Mark Christianson Content Project Managers: Kathryn D Wright, Bruce Gin, and Karen Jozefowicz Buyer: Jennifer Pickel Design: Debra Kubiak Content Licensing Specialists: Shawntel Schmitt and Shannon Manderscheid Cover Images: Cropped shot of young male skateboarder photographing feet on smartphone: © Cultura/Chad Springer/Getty Images; (bottom row) Vertu manufacturing/work stations and device assembly: Courtesy of Vertu; McDonnell Douglas DC-10-30F cargo aircraft taking on load: © Charles Thatcher/Getty Images; Store Manager assisting customer in phone store: © Echo/Getty Images Compositor: SPi Global Printer: R R Donnelley All credits appearing on page or at the end of the book are considered to be an extension of the copyright page Library of Congress Cataloging-in-Publication Data Names: Cachon, Gérard, author | Terwiesch, Christian, author Title: Operations management/Gerard Cachon, Christian Terwiesch Description: New York, NY : McGraw-Hill Education, [2017] Identifiers: LCCN 2015042363 | ISBN 9781259142208 (alk paper) Subjects: LCSH: Production management | Industrial management Classification: LCC TS155 C134 2017 | DDC 658.5—dc23 LC record available at http://lccn.loc.gov/2015042363 The Internet addresses listed in the text were accurate at the time of publication The inclusion of a website does not indicate an endorsement by the authors or McGraw-Hill Education, and McGraw-Hill Education does not guarantee the accuracy of the information presented at these sites mheducation.com/highered www.downloadslide.net DEDICATION To my core: Beth, Xavier, Quentin, Annick, and Isaac —Ge´rard To the Terwiesch family—in Germany, Switzerland, and the United States —Christian www.downloadslide.net About the Authors Gérard Cachon Gérard Cachon is the Fred R Sullivan Professor of Operations, Information, and Decisions and a professor of marketing at The Wharton School at the University of Pennsylvania Professor Cachon studies operations strategy with a focus on how new technologies transform competitive dynamics through novel business models He is the chair of the Operations, Information, and Decisions department; an INFORMS Fellow; a Fellow of the Manufacturing and Service Operations Management (MSOM) Society; a former president of MSOM; and a former editor-in-chief of Management Science and Manufacturing & Service Operations Management His articles have appeared in Harvard Business Review, Management Science, Manufacturing & Service Operations Management, Operations Research, Marketing Science, and the Quarterly Journal of Economics, among others At Wharton, he teaches the undergraduate course in operations management, and an MBA and executive MBA elective on operations strategy Before joining the Wharton School in July 2000, Professor Cachon was on the faculty at the Fuqua School of Business, Duke University He received a Ph.D from The Wharton School in 1995 He is a bike commuter (often alongside Christian) and enjoys photography, hiking, and scuba diving Christian Terwiesch Christian Terwiesch is the Andrew M Heller Professor at The Wharton School of the University of Pennsylvania He is a professor in Wharton’s Operations, Information, and Decisions department; is co-director of Penn’s Mack Institute for Innovation Management; and also holds a faculty appointment in Penn’s Perelman School of Medicine His research appears in many of the leading academic journals ranging from operations management journals such as Management Science, Production and Operations Management, Operations Research, and The Journal of Operations Management to medical journals such as The Journal of General Internal Medicine, Medical Care, Annals of Emergency Medicine, and The New England Journal of Medicine Most of Christian’s current work relates to using operations management principles to improve health care This includes the design of patient-centered care processes in the VA hospital system, studying the effects of emergency room crowding at Penn Medicine, and quantifying the benefits of patient portals and remote patient monitoring Beyond operations management, Christian is passionate about helping individuals and organizations to become more innovative Christian’s book Innovation Tournaments (Harvard Business School Press) proposes a novel, process-based approach to innovation that has led to innovation tournaments in organizations around the world Christian teaches MBA and executive classes at Wharton In 2012, he launched the first massive open online course (MOOC) in business on Coursera He also has been the host of a national radio show on Sirius XM’s Business Radio channel Christian holds a doctoral degree from INSEAD (Fontainebleau, France) and a diploma from the University of Mannheim (Germany) He is a cyclist and bike commuter and so, because his commute significantly overlaps the commute of Gérard, many of the topics in this book grew out of discussions that started on the bike After 15 years of Ironman racing, Christian is in the midst of a transition to the sport of rowing Unfortunately, this transition is much harder than predicted vi www.downloadslide.net Preface This introductory-level operations management title provides the foundations of operations management The book is inspired by our combined 30 years teaching undergraduate and MBA courses and our recent experience teaching thousands of students online via Coursera Seeing the need for a title different from our (highly successful) MBA textbook, we  developed this new book for undergraduate students and the general public interested in operations To engage this audience, we have focused our material on modern operations and big-picture operations Modern operations means teaching students the content they need in today’s world, not the world of 30 or 40 years ago As a result, “services” and “global” are incorporated throughout, rather than confined to dedicated chapters Manufacturing, of course, cannot be ignored, but again, the emphasis is on contemporary issues that are relevant and accessible to students For example, a Materials Requirement Planning (MRP) system is important for the functioning of a factory, but students no longer need to be able to replicate those calculations Instead, students should learn how to identify the bottleneck in a process and use the ideas from the Toyota Production System to improve performance And students should understand what contract manufacturing is and why it has grown so rapidly In sum, we want students to see how operations influence and explain their own experiences, such as the security queue at an airport, the quality of their custom sandwich, or the delay they experience to receive a medical test at a hospital Big-picture operations mean teaching students much more than how to math problems Instead, the emphasis is on the explicit linkages between operations analytics and the strategies organizations use for success For example, we want students to understand how to manage inventory, but, more importantly, they should understand why Amazon.com is able to provide an enormously broad assortment of products Students should be able to evaluate the waiting time in a doctor’s office, but also understand how assigning patients to specific physicians is likely to influence the service customers receive In other words, big-picture operations provide students with a new, broader perspective into the organizations and markets they interact with every day We firmly believe that operations management is as relevant for a student’s future career as any other topic taught in a business school New companies and business models are created around concepts from operations management Established organizations live or die based on their ability to manage their resources to match their supply to their demand One cannot truly understand how business works today without understanding operations management To be a bit colloquial, this is “neat stuff,” and because students will immediately see the importance of operations management, we hope and expect they will be engaged and excited to learn We have seen this happen with our own students and believe it can happen with any student vii Final PDF to printer www.downloadslide.net Acknowledgments This project is the culmination of our many years of learning and teaching operations management As such, we are grateful for the many, many individuals who have contributed directly and indirectly, in small and large ways, to our exploration and discovery of this wonderful field We begin with the thousands of students who we have taught in person and online It is through them that we see what inspires Along with our students, we thank our coteachers who have test piloted our material and provided valuable feedback: Morris Cohen, Marshall Fisher, Ruben Lobel, Simone Marinesi, Nicolas Reinecke, Sergei Savin, Bradley Staats, Xuanming Su, and Senthil Veeraraghavan We have benefited substantially from the following careful reviewers: Bernd Terwiesch took on the tedious job of proofreading early drafts of many chapters Danielle Graham carefully read through all page proofs, still finding more mistakes than we would like to admit We also thank Kohei Nakazato for double checking hundreds of test bank questions “Real operations” can only happen with “real” people We thank the following who matched supply with demand in practice and were willing to share their experiences with us: Jeff Salomon and his team (Interventional Radiology unit of the Pennsylvania Hospital System), Karl Ulrich (Novacruz), Allan Fromm (Anser), Cherry Chu and John Pope (O’Neill), Frederic Marie and John Grossman (Medtronic), Michael Mayer (Johnson&Johnson), and Brennan Mulligan (Timbuk2) From McGraw-Hill we thank our long-term friend Colin Kelley, who started us on this path and kept us motivated throughout, and the team of dedicated people who transformed our thoughts into something real: Christina Holt, Dolly Womack, Britney Hermsen, Doug Ruby, Kathryn Wright, Bruce Gin, and Debra Kubiak Finally, we thank our family members Their contributions cannot be measured, but are deeply felt Ge´rard Cachon   Christian Terwiesch We are grateful to the following professors for their insightful feedback, helpful suggestions, and constructive reviews of this text Stuart Abraham, New Jersey City University Khurrum Bhutta, Ohio University—Athens Greg Bier, University of Missouri—Columbia Rebecca Bryant, Texas Woman’s University Satya Chakravorty, Kennesaw State University Frank Chelko, Pennsylvania State University Tej Dhakar, Southern Hampshire University Michael Doto, University of Massachusetts—Boston Wedad Elmaghraby, University of Maryland Kamvar Farahbod, California State University—San Bernardino Gene Fliedner, Oakland University James Freeland, University of Virginia Phillip Fry, Boise State University Brian Gregory, Franklin University Roger Grinde, University of New Hampshire Haresh Gurnani, Wake Forest University Gajanan Hegde, University of Pittsburgh Michael Hewitt, Loyola University—Chicago Stephen Hill, University of North Carolina— Wilmington Zhimin Huang, Hofstra University Faizul Huq, Ohio University—Athens Doug Isanhart, University of Central Arkansas Thawatchai Jitpaiboon, Ball State University Peter Kelle, Louisiana State University—Baton Rouge Seung-Lae Kim, Drexel University Ron Klimberg, St Joseph’s University Mark Kosfeld., University of Wisconsin—Milwaukee John Kros, East Carolina University Dean Le Blanc, Milwaukee Area Technical College Matthew Lindsey, Stephen F Austin State University David Little, High Point University Alan Mackelprang, Georgia Southern University Douglas L Micklich, Illinois State University William Millhiser, Baruch College Ram Misra, Montclair State University viii cac42205_fm_i-xviii.indd viii 04/20/16 07:14 AM www.downloadslide.net Acknowledgments Adam Munson, University of Florida Steven Nadler, University of Central Arkansas John Nicholas, Loyola University—Chicago Debra Petrizzo, Franklin University William Petty, University of Alabama—Tuscaloosa Rajeev Sawhney, Western Illinois University Ruth Seiple, University of Cincinnati Don Sheldon, Binghamton University Eugene Simko, Monmouth University James E Skibo, Texas Woman’s University Randal Smith, Oregon State University James Stewart, University of Maryland University College ix Yang Sun, California State University—Sacramento Sue Sundar, University of Utah—Salt Lake City Lee Tangedahl, University of Montana Jeffrey Teich, New Mexico State University—Las Cruces Ahmad Vessal, California State University—Northridge Jerry Wei, University of Notre Dame Marilyn Whitney, University of California—Davis Marty Wilson, California State University—Sacramento Peter Zhang, Georgia State University Faye Zhu, Rowan University Zhiwei Zhu, University of Louisiana—Lafayette www.downloadslide.net Process Improvement LEARNING OBJECTIVES LO4-1 Compute the costs of direct labor, labor content, idle time, and average labor utilization LO4-4 Balance a process by reallocating work from one step to another LO4-2 Compute the takt time of a process and translate this to a target manpower LO4-5 Explain the benefits and limitations of specialization LO4-6 Evaluate the financial benefits of process improvements LO4-3 Find ways to improve the process efficiency by offloading the bottleneck CHAPTER OUTLINE Introduction 4.1 Measures of Process Efficiency 4.2 How to Choose a Staffing Level to Meet Demand 4.3 Off-Loading the Bottleneck 4.4 How to Balance a Process 4.5 The Pros and Cons of Specialization 4.6 Understanding the Financial Impact of Process Improvements Conclusion Introduction When Carl Benz built the first car with a combustion engine in 1886, he most likely did not care about the production process His focus was on the vehicle he produced His wife, Bertha Benz, was the key investor in this venture She was the first person to drive a car for a meaningful distance when she drove her husband’s invention from Mannheim, Germany, to the city of Pforzheim, which is roughly a 100 km drive This story would not be complete without mentioning that en route Bertha Benz ran out of fuel For obvious reasons, no gas stations existed along the way, so she purchased her fuel at a local pharmacy, which enabled her to successfully complete her journey back to Mannheim and secured her and her husband a place in automotive history Thirty years later, Henry Ford created the assembly line But Ford did not employ thousands of automotive experts akin to Carl Benz He hired unskilled labor off the street Each worker would only have to master a couple of seconds of work, so labor was cheap and plentiful Carl Benz spent months making one vehicle; Henry Ford built hundreds of vehicles in one day More vehicles at lower costs—that accomplishment secured Henry Ford a place in business history Henry Ford understood how it is possible to create operations that increased output in a Courtesy of Library of Congress Prints and Photographs Division [LC-USZ62-50219] given time period while also using relatively unskilled labor Obtaining high output at low costs is the key idea behind efficiency, the focus of the present chapter Here, we will start out by introducing a set of measures for efficiency Before diving into the details of efficiency, however, 67 www.downloadslide.net 68 Chapter Four  Process Improvement it is important to remind ourselves that efficiency is only one of multiple dimensions of operational performance and that businesses should not set efficiency as their only goal In fact, setting ethical considerations aside (and somewhat simplifying the world of finance), we can think of profit maximization as the primary goal for most corporations, where Profit = Flow rate × (Average price - Average cost) In the next sections, we discuss various forms of improving the efficiency of a process We then discuss how these improvement opportunities impact the profits derived from the operation We will this in the following steps: We start by developing a set of measures of efficiency After all, what you cannot measure, you cannot manage We then turn to finding the right number of employees for a given level of customer demand Having more process capacity than demand does not increase flow rate, yet it is expensive because we have to pay our employees This creates extra costs Having less process capacity than demand reduces flow rate, and thus reduces revenue We then will try to help the bottleneck resource to increase its capacity Because the process capacity is dictated by the resource with the smallest capacity, this will help us increase capacity and potentially the flow rate We then will attempt to divide up the work equally across the employees working in the process This avoids paying wages to labor that is underutilized, which, again, will help keep costs low and might lead to a higher flow rate The next level in improving efficiency is specialization To the extent that we can design the process flow in a way that our employees specialize in one or two activities, it is possible that we can reduce the time they take to carry out the activities, which would lead to more capacity Specialization is also often related to lower wages, which would lower our costs Finally, we will evaluate the impact of improvement opportunities on the profits of the operation Most importantly, we will see that even small efficiency improvements can have large financial benefits, especially when the process is presently capacity-constrained Given that efficiency drives the profits of an operation, managing an operation efficiently is a very valuable skill It does not matter if you work in a five-person startup or in a multibilliondollar enterprise, efficiency (and, of course, profits) matters everywhere in business It is also misleading to believe that efficiency is limited to manufacturing workers Hospitals aim to use their doctors and nurses efficiently Energy companies want to efficiently use their power plants Marketing executives want to efficiently utilize their salesforce And, as a society, we ought to make efficient use of our resources, be they water, land, or energy So, efficiency is key to many parts of our lives To illustrate the concepts of efficiency, we perform an analysis of a Subway restaurant And, in the spirit of Henry Ford’s auto plant, we will focus on the process flow that employs specialProfit maximization The objective of an enterprise—to maximize the difference between revenue and costs ization Instead of one person taking care of all the activities, from greeting the customer to running the cash register, we consider a process where multiple employees serve the customer, each worker specializing in a different set of activities www.downloadslide.net 69 Chapter Four  Process Improvement 4.1  Measures of Process Efficiency Subway provides custom-made sandwiches to its customers Consider the three-step process of serving a customer depicted in Figure 4.1 It takes the first employee (“station”) 37 seconds to greet the customer and perform all activities up to adding the cheese onto the sandwich From there, employee takes over and performs all activities from placing the onions to wrapping and bagging, which, taken together, takes 46 seconds per customer Finally, station offers a fresh value meal and a cookie and then rings up the purchase on the register, which takes together 37 seconds per customer In the following calculations, we assume that these times are exact; that is, that it will take an employee exactly that amount of time to carry out an activity Moreover, we assume that every customer wants all the ingredients Arguably, these are strong assumptions, but they make for a good starting point in our analysis Depending on what dictionary you consult, you will find the word “efficiency” defined as “acting directly to produce an effect,” “acting or producing effectively with a minimum of waste, expense, or unnecessary effort,” and “obtaining a high ratio of output to input.” Whatever definition you want to pick, all of them share what in business is often referred to as a “big bang for the buck.” So, in the context of process analysis, we define a process as efficient if it is able to achieve a high flow rate with few resources In the case of our Subway analysis, our focus is primarily on the employees (the labor) as the key resource (as opposed to the oven or the real estate) And so rather than talking about efficiency in general, we will focus on labor efficiency A first measure of labor efficiency is the labor cost associated with serving one customer We call this measure the costs of direct labor We compute the costs of direct labor as Wages per unit of time _ ​Costs of direct labor =  ​        ​​ Flow rate Station Station Station Efficiency A process is efficient if it is able to achieve a high flow rate with few resources Costs of direct labor The labor cost associated with serving one customer, which is the total wages paid per unit of time divided by the flow rate LO4-1  Compute the costs of direct labor, labor content, idle time, and average labor utilization Figure 4.1  Process flow for serving one customer, assuming three employees Station Station 37 sec/customer Station Processing time Waiting customers Activity Greet customer Take order Get bread Cut bread Meat Cheese Onions Lettuce Tomatoes Cucumbers Pickles Green peppers Black olives Hot peppers Place condiments Wrap and bag Offer fresh value meal Offer cookies Ring up Register Total 46 sec/customer 37 sec/customer Activity Time [sec/customer] 12 3 4 13 14 20 120 www.downloadslide.net 70 Chapter Four  Process Improvement How many customers will the process shown in Figure 4.1 serve? The first station has a capacity of _ ​​  1  ​​ customer/second, the second one of _ ​​ 1  ​​ , and the third one of _ ​​ 1  ​​.  Hence, the 37 37 46 second station has the lowest capacity and is the bottleneck Assuming sufficient demand, we are thus making a sandwich at a rate of _ ​​ 1  ​​ customer/second × 3600 seconds/hour = 78.3 46 customers/hour Given that we presently have three employees in our process and assuming that we pay them $12 per hour, we get × 12 $ / h _ ​Costs of direct labor =  ​        ​  = 0.46 $ / customer​ 78.3 customers / h If we think about output and input (or the bang and the buck, if you prefer), we could casually say that the wages we pay our employees are an input into the operation and the number of sandwiches that the labor produces constitutes an output Thus, the definition of the costs of direct labor looks at the ratio of inputs to outputs Note that this is slightly different from the dictionary definition that called for computing the ratio of output to input To state the obvious, high costs of direct labor are associated with a low labor efficiency and our measure is just the reciprocal of the dictionary definition But the ratio of wages and flow rate are not the only output-to-input calculations we can perform For example, we might think about the amount of work that goes into serving one customer as the input into the process and turning that customer into a served customer as the output With this in mind, we define a second efficiency measure as the labor content: ​Labor content = Sum of the processing times involving labor​ The labor content measures how much work is required in serving one flow unit As with the costs of direct labor, holding everything else constant, a high number is arguably less desirable than a low number and so we are again looking at a ratio of input to output In our Subway example, we get  seconds  ​Labor content = 37 ​   ​  + 46 ​  seconds   ​  + 37 ​  seconds   ​  customer customer customer seconds  = 120 ​   ​​  customer Yet another measure of labor efficiency can be derived based on the utilization ­levels of the employees in the process Assuming sufficient demand, we determined a flow  rate of 78.3 ­customers/hour The three stations have a capacity of _ ​​ 1  ​​ customer/second, _ ​​  1  ​​ customer/ 37 46 second, and _ ​​  1  ​​ customer/second, which can be translated to hourly capacity l­evels of 97.3, 37 78.3, and 97.3, respectively Because the utilization of a resource is the ratio between its flow rate and its capacity, we obtain utilization levels of 80.4 percent, 100 percent, and 80.4 ­percent So we can compute the average labor utilization in a process as Labor content The amount of work that goes into serving one customer (or, more generally, one flow unit), which is the sum of the processing times involving labor Average labor utilization  The average utilization across resources Idle time The amount of time per flow unit for which a resource is paid but is not actually working Average labor utilization = Average utilization across employees       ​       ​ =​ Average(80.4%, 100%, 80.4%)   ​ ​ ​ = 87.0% The average labor utilization is a measure of efficiency—we have to pay our employees whether they are working or not Thus, an unutilized worker creates an unnecessary expense and so an efficient process is one in which the average labor utilization is as high as possible We refer to the time we pay a resource even though the resource is not working as idle time We can compute the idle time for each resource as Idle time(Resource i) = Cycle time - Processing time(Resource i) The idle time measures how long the resource is idle for each flow unit it serves, which is why the idle time is expressed in seconds/customer (or, more generally, in units of time per flow www.downloadslide.net Chapter Four  Process Improvement 71 unit) The cycle time tells us how long we have to wait between completing the order of two consecutive customers The cycle time is 1/Flow rate, in this case 46 seconds/customer So, we get seconds  ​Idle time(Station 1) = 46 - 37 = 9 ​   ​  customer seconds  Idle time(Station 2) = 46 - 46 = 0 ​   ​  customer seconds  Idle time(Station 3) = 46 - 37 = 9 ​   ​​  customer We can also add up the idle time across the employees, which gives us the total idle time in the process: ​Total idle time = Sum of idle time across resources seconds  = + + 9 = 18 ​   ​​  customer So, for every customer that we serve, we incur (and pay for) 18 seconds of idle time Is this a lot? This question is difficult to answer Eighteen seconds of idle time per unit seems like a small number in the assembly of a car, in which hundreds of people and multiple hours of labor content per vehicle are involved But for a sandwich? To evaluate the magnitude of the idle time, it is useful to compare the idle time with the labor content Recall that we had defined the labor content as the sum of the processing times in the process—in our case, 120 seconds/customer Now, compare the labor content (which we can think of as the productive time of our resources) relative to the labor content plus the idle time (which we can think of as the total time that we have to pay our employees for each unit because we pay them if they are working or if they are idle) We get 120 seconds/customer Labor content ​​         ​ = ​  _           ​ = 87.0%​ Labor content + Idle time 120 seconds/customer +18 seconds/customer Note that this is exactly the same value as the average labor utilization This is not a coincidence In both calculations, we compare the idle time to the time that the employees actually work When we computed the average across the utilizations, we took the perspective of the employees When we looked at the labor content relative to the labor content + idle time, we took the perspective of the flow unit Either way, we can generalize and write Labor content ​Average labor utilization = _ ​        ​​ Labor content + Idle time There exists a third way to obtain the average labor utilization, which is based on the cycle time: Labor content / Number of employees ​Average labor utilization = ​         ​​ Cycle time Let’s just try it out: Labor content / number of employees _ ​​         ​ = ​  120/3  ​   = _ ​  40 ​ = 87.0 percent​ Cycle time 46 46 The intuition behind this third way of computing the average labor utilization is as follows In a perfectly balanced process (a process that has an average labor utilization of 100 percent), the labor content is divided up equally across the number of employees In the case of three Total idle time The amount of idle time per flow unit added up across all resources www.downloadslide.net 72 Chapter Four  Process Improvement employees, each employee gets a third of the labor content as his or her processing time This allows them to produce at a cycle time that equals their (common) processing time So, three different ways of computing the average labor utilization all get the same result: which one should you pick? Pick whichever is easiest for you The benefit of the last measure is that it also works in processes where you have different staffing levels across resources (for example, three employees at station 7, but only one employee at station 2) Check Your Understanding 4.1 Question:  Consider the example of the three-step process The first step takes 20 minutes/ unit, the second step takes 10 minutes/unit, and the third step takes 15 minutes/unit Each step is staffed by one worker What is the labor content of the process? Answer:  The labor content is 20 + 10 + 15 = 45 minutes/unit Question:  What is the total idle time of the process assuming unlimited demand? Answer:  We first have to establish that the first step is the bottleneck It has the lowest capacity (1/20 unit/minute) This gives us a process capacity of 1/20 unit/minute Given that we have unlimited demand, we also get a flow rate of 1/20 unit/minute and a cycle time of 1/Flow rate = 20 minutes/unit With this, we can compute the idle times for the three steps as minutes ​Idle time(Step 1) = 20 - 20 = ​   ​     unit minutes Idle time(Step 2) = 20 - 10 = 10 _ ​   ​     unit minutes Idle time(Step 3) = 20 - 15 = ​   ​     unit minutes So, the total idle time is + 10 + = 15 _ ​   ​​    unit Question:  What is the average labor utilization? Answer:  We compute the average labor utilization as Labor content  ​Average labor utilization =    ​      ​ (Labor content + Idle time) 45 = ​       ​ = 0.75​ (45 + 15) Note that this is the same as Labor content /Number of employees ​Average labor utilization =      ​       ​ Cycle time 45/3 _ 15 = ​   ​ = ​    ​ = 0.75​ 20 20 Question:  What is the cost of direct labor, assuming that each of the three employees gets paid $15 per hour units Answer:  We already computed that the flow rate is _ ​​     ​  unit/minute = ​   ​​  Given that we hour 20 have three employees, we have to pay × $15 per hour = $45 per hour, giving us $45 per hour ​Cost of direct labor = _ ​       ​ = $15 per unit​ 3 units per hour www.downloadslide.net Chapter Four  Process Improvement 73 4.2  How to Choose a Staffing Level to Meet Demand Matching supply with demand almost always starts with taking the demand rate as given and attempting to staff a sufficient number of resources in order to meet that demand So, let us assume that our current demand in the Subway restaurant is given and it is 100 customers per hour To serve 100 customers per hour, we have to produce at one customer every 36 seconds This is called the takt time of the process LO4-2  Compute the takt time of a process and translate this to a target manpower 1   ​  1  ​Takt time = _ ​  = ​      ​ Demand rate (100 customers/hour) seconds seconds        = ​  0.01 hour ​  × 3600 ​   ​    = 36 ​  _  ​​  customer customer hour Note that takt time is a measure that is entirely driven by demand; it is our goal to design a process flow that meets exactly this demand rate Takt is a German word—the takt of the music tells the musician at which tempo (speed) to play a particular piece The tempo of a musical piece is typically expressed in beats per minute—or, to be more exact, the number of quarter notes to be played per minute (just for completeness, in music, tempos lower than 100 beats per minute are typically considered slow, while tempos above 100 beats are considered fast) The fact that takt is a musical term is appropriate Just as how the musician should not play the notes in a musical piece at her own discretion, a process should not be operating at the discretionary flow of the resources Instead, the flow should be happening at the rate of demand In the previous calculations of takt time, we assumed that demand was expressed in the form of a demand rate—that is, a required quantity per unit of time Sometimes, however, demand might be expressed as a quantity, say of 100 units In that case, we find the takt time as follows: Available time ​Takt time = _ ​       ​​ Required quantity where Available time measures the amount of time we have available in the process to produce the quantity For example, if we work 8-hour work shifts and we want to produce 120 widgets, we have to produce at a takt time of × 60 minutes Available time  ​ = ​  _ ​Takt time = _ ​              ​ Required quantity 120 widgets​ minutes ​​  = 4 ​  widget Note that this definition of takt time is a generalization of the previous definition of takt time = 1/Demand rate in which we set the Available time equal to one unit of time It is important to understand the difference between takt time and cycle time Recall that cycle time is 1/Flow rate (and that the flow rate is the minimum of demand and process capacity) So, cycle time depends on process capacity; takt time does not The only thing that drives takt time is the demand rate Cycle time, as 1/Flow rate, measures the current reality of the process flow In the spirit of matching supply with demand, our goal is to have a cycle time that is as close to the takt time as possible If customers order a sandwich every 36 seconds, our goal ought to be to produce a sandwich every 36 seconds Instead of matching our cycle time to our takt time, we can also think about matching our capacity to the demand rate These two approaches are entirely equivalent In both cases, we will adjust the staffing levels If we are currently capacity-constrained, we will add more employees More employees mean a higher capacity More capacity in turn will lead to a higher flow rate (because we are capacity-constrained) and that translates to a shorter cycle time If we are currently demand-constrained, we will aim to reduce the number of employees Takt time The ratio between the time available and the quantity that has to be produced to serve demand www.downloadslide.net 74 Chapter Four  Process Improvement Fewer employees mean a lower capacity and thus a capacity that is closer to the demand rate Fewer employees will not hurt our cycle time, because our flow rate is currently driven by the demand rate In our Subway example, facing a demand rate of 100 customers per hour, our three-step process (with processing times of 37, 46, and 37 seconds/unit, respectively) is capacity-­ constrained and the flow rate is 78.3 customers per hour Thus, we leave 21.7 customers per hour not served (they go home hungry) But how many employees we need to add? Assume for the moment that every employee would be fully utilized (that is a big IF, but more on this later) We can compute the required staffing level using the target manpower formula: Labor Content ​Target Manpower = ​     ​​  Takt Time As computed above, we have 1   ​  1  ​Takt time = _ ​  = _ ​      ​ Demand rate 100 customers/hour hour  ​  sec  ​ = 36 ​  seconds  = 0.01 ​  × 3600 ​   ​​  customer customer hour So, our target manpower is ​Target Manpower =    ​  Labor Content  ​   Takt Time 120 seconds/customer ​ = 3.333​ ​= ​       36 seconds / customer Assuming that all our employees are fully utilized, we need 3.33 employees Because it is hard to hire a fraction of an employee, we have to round this up to four employees For ease of notation, we refer to the target manpower as 3.333 and the staffing level as In other words, the target manpower can be a decimal; the staffing level, however, can only be an integer Should you always round up to get the staffing level? What about the case in which the target manpower is 3.0001? The answers to these questions depend on your goal If your goal is to definitely meet demand in its entirety (i.e., not turn a single customer away), you should round up If your goal is to maximize profits, it is a little more complicated In this case, you should the calculations with a staffing level obtained from rounding down, and then with a staffing level obtained from rounding up; you then compare the profits (as we at the end of the chapter) and determine which of these two options yields more profits Now, assume that demand goes up to 160 sandwiches per hour This is a very high demand rate, even for large Subway restaurants A higher demand rate means that we have a lower takt time The tempo of the music has picked up—we need to operate at a faster cycle time to keep up with the takt of demand How many employees we need now? Again, we start by computing the takt time: 1   ​  1  ​Takt time = _ ​  = _ ​     ​ Demand rate 160 customers / hour hour  ​  sec  ​ = 22.5 ​  seconds  = 0.00625 ​  × 3600 ​   ​​  customer customer hour Target manpower The ratio between the labor content and the takt time determines the minimum number of resources required to meet demand Note that this minimum does not have to be an integer number and that it assumes all resources are perfectly utilized We can then use the target manpower formula to find the required staffing level: Labor content ​Target manpower = ​     ​   Takt time 120 seconds/customer  ​ = 5.333​ ​​= _ ​       22.5 seconds/customer www.downloadslide.net 75 Chapter Four  Process Improvement So, more demand means a shorter takt time, and a shorter takt time requires more employees to handle the same amount of labor content As explained earlier, instead of deriving the staffing level based on takt time, we might also ask ourselves: How many employees would it take to obtain a capacity of 160 customers per hour (which is the demand rate)? Given an unknown number of employees, m, we know that the capacity of a resource with m employees is 3600 × m m  m  ​Capacity = _    ​   ​ =    ​   ​ = ​  _        ​​ Processing time 120 seconds/customer 120 customers / hour If we substitute Capacity = 160 units/hour, we get the equation 3600 × m customers ​160 ​  _  ​    = ​         ​ hour 120 customers/hour ⇔ 160 = 30 × m  ⇔  m = 5.333​ So, the solution comes out exactly the same and we leave it to you if you want to find the target manpower level by matching capacity to flow rate or by matching cycle time to takt time The following calculations, as you will see, are easier to follow when working with takt time, and so we keep our focus on takt time In most businesses, demand varies, be it by the hour of the day, the day of the week, or the time of the year We typically cannot adjust our staffing level for every little change in demand and so the first thing that operations managers in practice for planning purposes is to level the demand; that is, set an expected demand rate for a given period of time (say one hour at a Subway restaurant) Then, we turn that leveled demand into a takt time (remember: high demand means a short takt time) and then translate the takt time into a staffing level That way, we always staff to demand Figure 4.2 captures how an operation can translate a time-varying demand and adjust the staffing over time Leveling the demand Setting an expected demand rate for a given period of time so that one can look for an appropriate staffing plan for that time period Check Your Understanding 4.2 Question:  A large manufacturer of laptop computers operates two eight-hour shifts; thus, it has 16 hours available for production The goal of the manufacturer is to produce 480 ­computers a day What takt time does the manufacturer have to set? Answer:  We know that Available time ​Takt time =    ​    ​ Required quantity/hour 16 hours/day = ​       ​ 480 computers/day hour = 0.0333 ​    ​  computer minutes = 2 ​     ​​  computer Question:  Now assume it takes 20 minutes of labor to assemble a computer What would be the target manpower? Answer:  The target manpower would be 20 minutes/computer Labor Content _ ​Target Manpower =    ​      ​ = ​       ​ = 10​ Takt Time minutes/computer © Sot/Digital Vision /Getty Images/RF www.downloadslide.net 76 Chapter Four  Process Improvement Figure 4.2 Actual Demand Demand leveling and its impact on staffing Volume 160 80 Time Leveled Demand Volume 160 80 80 Takt time Takt 45 45 22.5 Resource planning Staffing level 3 Given our takt time of 22.5 seconds per customer, let us now try to create a process that is able to work according to that beat of demand If we assume for now that each activity is only carried out by one and only one worker, then we have to keep all processing times at 22.5 seconds/customer or below In other words, the activities carried out by one employee can together not take longer than 22.5 seconds/customer This simply reflects that to achieve a flow 1   ​​ customer per second), each resource in the rate of the process of 160 customers per hour (​​  22.5 1   ​​ customer per second) process must have a capacity of at least 160 customers per hour (​​  22.5 Further, assume that the activities in the process recipe have to be carried out in the sequence in which they appear in Figure 4.1 Thus, as we are now deciding which employee will carry out which activity, we can simply start with the first employee (resource) and keep on adding activities to the resource as long as her processing time remains at or under 22.5 seconds/unit Or, put differently, we continue to load activities to the resource until the resulting processing time of the resource exceeds 22.5 seconds/customer and then take the last activity added and move it to the next resource That sounds complicated, but really is not Let’s just the calculations: • The first employee in the process can greet the customer (4 seconds/customer) Because that still is well below the takt time, we give the employee additional work and have her take the order (5 seconds/customer) The resulting processing time would now be + = seconds/customer, still well below the 22.5 target So, we keep on www.downloadslide.net 77 Chapter Four  Process Improvement • • • • • • adding work, including getting the bread (4 seconds/customer) and cutting the bread (3 seconds/customer) That now gets us to + + + = 16 seconds/customer The next activity is dealing with the meat, which takes 12 seconds/customer of work Because 16 + 12 = 28 > 22.5, this is too much work for the employee to operate on a 22.5-seconds-per-customer cycle So we leave the first employee in charge of greeting, taking the order, getting the bread, and cutting the bread and get a resulting processing time of 16 seconds/customer The second employee gets to deal with the meat, the activity that would have moved the processing time above 22.5 seconds per customer for the first employee We can also put employee in charge of the cheese (9 seconds/customer) because 12 + = 21 < 22.5 But that’s it—the next activity (the onions, seconds/customer) needs to be carried out by employee For employee 3, we start with the onions (3 seconds/customer) and can add lettuce (3 seconds/customer), tomatoes (4 seconds/customer), cucumbers (5 seconds/customer), and pickles (4 seconds/customer), yielding a processing time of + + + + = 19 seconds/customer Employee then can deal with the green peppers, black olives, and hot peppers, with activity times of 4, 3, and seconds per customer, respectively Because the resulting processing time of + + = seconds/customer is still below our 22.5 seconds/ customer target, we can also have her add the condiments (5 seconds/customer), getting the processing time to 14 seconds/customer Employee then can wrap and bag and offer the fresh value meal, getting a processing time of 13 + = 16 seconds/customer Employee can offer cookies but will not be able to ring up the sale on the register, leaving her with 14 seconds/customer And, finally, employee will be in charge of the cash register, yielding a 20-secondper-customer processing time The resulting allocation of activities to the seven employees is shown in Figure 4.3 Each box in the figure corresponds to one activity, with the height of the boxes capturing the duration of the activity So, the cumulative height across boxes that are stacked on each other corresponds to the processing time In addition to the seven processing times, the figure also Figure 4.3  A Subway sandwich line with seven employees Takt time 22.5 seconds/unit Cheese (9) Cut bread (3) Get bread (4) Take order (5) Pickles (4) Offer fresh value meal (3) Cucumbers (5) Place condiments (5) Tomatoes (4) Hot peppers (2) Lettuce (3) Black olives (3) Greet customer (4) Meat (12) Onions (3) Green peppers (4) Wrap and bag (13) Offer cookies (14) Ring up on register (20) Employee Employee Employee Employee Employee Employee Employee www.downloadslide.net 78 Chapter Four  Process Improvement shows our takt time We observe that all resources have processing times below the takt time, indicating that the process will be able to operate at the rate of demand But wait a minute! Why are we talking about seven employees? Didn’t we just compute that it would take 5.333 employees to achieve the takt time of 22.5 seconds/customer? Weren’t we “generous” by rounding up to a staffing level of six employees? And now we find that we need seven employees That does not look like efficiency at all But where is the inefficiency here? The answer is quite simply the idle time The target manpower calculation implicitly assumes that all resources are utilized at 100 percent—that is, there is no idle time in the process However, this is typically not the case And, as we will see in a moment, we have more idle time in the process compared to when we had three employees make the sandwich To see this, let us perform a careful process analysis of our seven-step process We compute the capacity levels of each resource, which will get us the bottleneck We then can compute flow rate as the minimum between demand and process capacity From there, we can get the utilization levels and the cycle time The resulting analysis is shown in Figure 4.4 We see that with a processing time of 21 seconds/customer and hence a capacity of _ ​​    ​​ customer/second, station is the bottleneck Further, we observe that the process is 21 ­presently demand-constrained, given that it has a process capacity of 171.4 customers/hour and a demand rate of 160 customers/hour This results in a flow rate of 160 customers/hour and a cycle time of one customer every 22.5 seconds As was done earlier in this chapter, we can compute the idle time for each resource as ​Idle time(Resource i) = Cycle time - Processing time(Resource i)​ So, we get ​Idle time(Station 1) = 22.5 - 16 = 6.5 ​  seconds   ​  customer Idle time(Station 2) = 22.5 - 21 = 1.5 ​  seconds   ​  customer Idle time(Station 3) = 22.5 - 19 = 3.5 ​  seconds   ​  customer Idle time(Station 4) = 22.5 - 14 = 8.5 ​  seconds   ​  customer Idle time(Station 5) = 22.5 - 16 = 6.5 ​  seconds   ​  customer Figure 4.4  Process analysis of the seven-station subway line Waiting Station Customers Station Station Station Station Station Station Processing time 16 sec/unit 21 sec/unit 19 sec/unit 14 sec/unit 16 sec/unit 14 sec/unit 20 sec/unit Capacity (per second) Capacity (per hour) unit/sec 16 225 units/h unit/sec 21 171.4 units/h unit/sec 19 189.5 units/h unit/sec 14 257.1 units/h unit/sec 16 225 units/h unit/sec 14 257.1 units/h unit/sec 20 180 units/h Process capacity Minimum{225 units/h, 171.4 units/h, 189.5 units/h, 257.1 units/h, 225 units/h, 257.1 units/h, 180 units/h} = 171.4 units/h Bottleneck? No Yes Flow rate Utilization Cycle time No No No No No 62.2% 88.9% Minimum{171.4 units/h, Demand} = 160 units/h 71.1% 93.3% 84.4% 62.2% 3600 sec/h = 22.5 sec/unit 160 units/h 71.1% www.downloadslide.net Chapter Four  Process Improvement Idle time(Station 6) = 22.5 - 14 = 8.5 ​  seconds   ​  customer Idle time(Station 7) = 22.5 - 20 = 2.5 ​  seconds   ​​  customer We add up the idle time across the employees, which gives us the total idle time in the process: ​Total idle time = Sum of Idle time across resources = 6.5 + 1.5 + 3.5 + 8.5 + 6.5 + 8.5 + 2.5 = 37.5 ​  seconds   ​​  customer So, for every customer that we serve, we incur (and pay for) 37.5 seconds of idle time Is this a lot? As done before, we compare the idle time with the labor content, a measure that we previously introduced as the average labor utilization: Labor content _ ​ Average labor utilization =    ​       ​ (Labor content + Idle time) 120 seconds/customer =     ​      ​ (120 + 37.5) seconds/customer = 76.2 percent​ Note that this number is lower than the process with three employees, where we obtained an average labor utilization of 87.0 percent This decrease in efficiency is also reflected in the costs of direct labor: Wages per unit of time _ ​Costs of direct labor =    ​      ​​ Flow rate With our three-person process, our costs of direct labor were $0.46 per customer With the seven employees, we now get × 12 $/h ​Costs of direct labor = ​  _       ​ 160 customers/ah $ = 0.525 ​     ​​  customer So, by running the process faster, we have actually increased our costs But why is this? To identify the source of the cost increase, consider again the average labor utilization With the three-employee process, that measure was 87.0 percent; now, with seven employees, it is down to 76.2 percent There are two reasons for this: • First, with a demand of 100 customers/hour and a process capacity of 78 customers/ hour, we were capacity-constrained That is, our bottleneck resource (employee 2) was operating at 100 percent utilization Now, in contrast, even our bottleneck resource (again, employee 2, though now with a different set of activities and a shorter processing time) is only running at a 93.3 percent utilization Thus, we have slack capacity relative to demand • Second, balancing the process becomes harder as we spread out the activities over more resources Consider the extreme case of having one worker being in charge of all activities Assuming sufficient demand, that resource would have an average labor utilization of 100 percent No surprise, a process with one employee is perfectly balanced But balancing the process becomes harder as we add more employees to the process and we spread out the activities over more and more resources But, again, the goal of an operation is not to maximize the labor utilization By adding more employees, we have suffered some efficiency loss (a couple of percentage points of 79 www.downloadslide.net 80 Chapter Four  Process Improvement labor utilization, a couple of pennies of costs of direct labor), but, in contrast, our flow rate has gone up by a lot Because a Subway sandwich sells for around $6 per customer and the material costs of a sandwich are substantially smaller than that, we most likely have increased our profits 4.3  Off-Loading the Bottleneck © Mario De Biaso/Digital Stock/Royalty-Free/Corbis/RF Herbie A fictitious character in Eli Goldratt’s book The Goal Herbie is the slowest hiker in a troop of Boy Scouts He thus holds up the troop in the same way a bottleneck holds up a process In the previous section, we saw how to choose an appropriate staffing level given a certain amount of demand Choosing the right staffing level makes sure that we don’t have too much capacity (leading to idle time and high costs of direct labor) but still serve demand (leading to no lost revenues from turning customers away) We now consider other forms in which we can improve the process to make it more efficient In 1984, Eli Goldratt wrote a highly successful book entitled The Goal. It’s hard to believe, but the book is an operations management textbook written as a novel and has sold millions of copies The hero of the book is a fictional plant manager by the name of Alex who discovers the principles of process analysis Alex experiences his epiphany when he takes a group of Boy Scouts for a hike in the forest The group takes a single-file path and starts out close to each other But the longer they march, the more the group spreads out The boys in the front tend to be the fast ones and so they walk at their own rate, leaving the others behind One chubby Scout by the name of Herbie is holding the group back He has hundreds of meters of empty trail in front of him and a group of impatient Boy Scouts behind him (Scouts who, due to the narrowness of the trail, cannot overtake Herbie) Alex realizes that this hike has a lot in common with his production process He associates the speed with which the entire group is moving with the flow rate in the plant Just as the plant cannot produce faster than the bottleneck, the group of Boy Scouts cannot walk faster than Herbie The amount of empty hiking path between the kids, Alex recognized, is similar to inventory in a plant where the inventory piles up in front of the bottleneck Since the publication of the book, many operations experts have referred to the bottleneck as Herbie The example of the Boy Scouts reminds us that whatever process improvement we might consider, we have to start with the bottleneck It might be possible to reduce the time required to put the meat on the sandwich (12 seconds/customer) For example, we could www.downloadslide.net already organize the meat on a little plastic foil, ready to be put on the sandwich, which might reduce the time of this activity to seconds/customer However, this improvement in and by itself will be of no value Employee is not the bottleneck, and so all that will be accomplished with this “improvement” is to increase the amount of idle time for the employee Such an improvement would be similar to making one of the fast Boy Scouts walk even faster and then having to wait for Herbie at the next campsite So, any process improvement starts by looking at the bottleneck In the case of Herbie, Alex soon found out that his ability to walk as fast as the others was not just constrained by his physiological limitations, but also by Herbie carrying a rather large and heavy backpack with plenty of snacks inside To improve the hiking speed of his group, Alex then takes Herbie’s backpack and spreads the content among the faster Scouts Yes, those are now slowed down, but Herbie is able to walk faster and that is all that counts In general, we refer to the improvement strategy of moving work away from the bottleneck step as off-loading the bottleneck Off-loading the bottleneck can take multiple forms: • 81 Chapter Four  Process Improvement Reassign activities to other resources with more capacity, an improvement strategy that we refer to as line balancing and discuss in the next section • Automating some of the activities consuming time at the bottleneck by using technology; for example, we might be able to automate the dispensing of the wrapping paper, which has the potential of reducing the processing time for employee • Outsourcing some of the activities consuming time at the bottleneck Imagine it would be possible to put the condiments in the bag at the beginning of the shift or to even purchase bags that are already “loaded” with condiments This would reduce the time for employee 2, which, given that employee is the bottleneck, would increase the capacity of the process LO4-3  Find ways to improve the process efficiency by off-loading the bottleneck In addition to off-loading the bottleneck, we can increase the process capacity by adding more employees If you add just one employee (or, in general, one unit of capacity), where would you put him or her? The answer is “to the bottleneck” because that is what constrains the process—the chain is only as strong as its weakest link Adding an employee to the bottleneck is like boosting the fitness of Herbie. . .  In addition to adding more employees, we can also boost the bottleneck capacity by either having the bottleneck resource work longer hours (overtime) or assigning a more skilled employee to the bottleneck resource, assuming that this employee would have a shorter processing time 4.4  How to Balance a Process In general, we balance a process by allocating the activities that need to be carried out in the process across the process resources as evenly as possible Because most of these tools have their roots in the management of assembly lines, we speak of line balancing rather than process balancing In practice, two cases of such balancing have to be distinguished: LO4-4  Balance a process by reallocating work from one step to another • Balancing for a fixed sequence of activities: In this case, we have to carry out the activities in a given order, one activity is the first, one the second, and so on This corresponds to what we did as we moved from the three-employee line to the sevenemployee line We first compute takt time and then keep on adding activities to a resource just so long as the resulting processing time of that resource stays under the takt time We then proceed to the next resource, until all activities are assigned • Balancing for activities with no fixed sequence: It is hard to wrap a sandwich before it is made But could you put hot peppers on a sandwich before dealing with the pickles? Probably you could In some cases, you can reshuffle the activities in your process This gives you an extra bit of flexibility, which makes assigning activities to resources in a balanced way easier, and thus should lead to a higher average labor utilization and lower costs of direct labor Line balancing The act of allocating the activities that need to be carried out in the process across the process resources as evenly as possible so that all resources have a comparable utilization level ... (min) ? ?1  7:35   8:50   75  2  7:45 10 :05 14 0  3  8 :10 10 :10 12 0  4  9:30 11 :15 10 5  5 10 :15 10 :30  ? ?15  6 10 :30 13 :35 18 5  7 11 :05 13 :15 13 0  8 12 :35 15 :05 15 0  9 14 :30 18 :10 220 10 14 :35 15 :45... achieve cac42205_ch06 _13 9 -17 3.indd 17 1 Focus on Process Analysis cac42205_ch06 _13 9 -17 3.indd 16 8 11 /23 /15 06:45 PM Written for the Connect Platform 11 /23 /15 06:45 PM Operations Management has been... Formulas  310 Conceptual Questions  310 Solved Example Problems  311 Problems and Applications  313 Case: Linking Turns to Gross Margin  315 11 Supply Chain Management? ?? 316 Introduction  316 Supply